---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\fsu2010\formateur\replication\stata_mxl.log
  log type:  text
 opened on:   5 Nov 2010, 16:19:10

. #delimit ;
delimiter now ;
. set mem 30m;

Current memory allocation

                    current                                 memory usage
    settable          value     description                 (1M = 1024k)
    --------------------------------------------------------------------
    set maxvar         5000     max. variables allowed           1.947M
    set memory           30M    max. data space                 30.000M
    set matsize         400     max. RHS vars in models          1.254M
                                                            -----------
                                                                33.201M

. clear matrix;

. clear mata;

. set more off;

. *     ***************************************************************** *;
. *     ***************************************************************** *;
. *       File-Name:  stata_mxl.do                                        *;
. *       Date:       November 5, 2010                                    *;
. *       Author:     GG/MG                                               *;
. *       Purpose:    Produce Tables 1 and 2 in AJPS paper                *;
. *       Input File:     europe.dta                                      *;
. *       Output File:    stata_mxl.log                                   *;
. *       Data Output:    none                                            *;
. *       Previous file:  europe.dta                                      *;
. *       Machine:        Matt's desktop                                  *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. use "C:\fsu2010\formateur\replication\europe.dta", clear;

. *     ****************************************************************  *;
. *       Produce the results shown in Tables 1 and 2.  The LR tests      *;
. *       compare each of the 6 models to a model where we only include   *;
. *       largest party as an independent variable.                       *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *       Model 1: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east,  
>     rand (largestparty party_seatshare) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -334.08276  (not concave)
Iteration 1:   log likelihood = -304.04279  
Iteration 2:   log likelihood =  -303.2049  
Iteration 3:   log likelihood = -301.76234  (not concave)
Iteration 4:   log likelihood = -301.49878  
Iteration 5:   log likelihood = -300.49577  
Iteration 6:   log likelihood = -300.46478  
Iteration 7:   log likelihood =  -300.4633  
Iteration 8:   log likelihood =  -300.4633  

Mixed logit model                                 Number of obs   =       2412
                                                  LR chi2(2)      =       8.00
Log likelihood =  -300.4633                       Prob > chi2     =     0.0183

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -.0772611   .8494944    -0.09   0.928    -1.742239    1.587717
party_se~ast |   .0996268   .0634406     1.57   0.116    -.0247145    .2239681
largestparty |   1.027553   .5402949     1.90   0.057    -.0314061    2.086511
party_seat~e |   .1369203   .0207696     6.59   0.000     .0962126     .177628
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.817642   1.078137    -2.61   0.009    -4.930752   -.7045326
party_seat~e |  -.0088345    .055768    -0.16   0.874    -.1181377    .1004687
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. mixlogit choice largestparty_east party_seatshare_east,   
>     rand (largestparty party_seatshare) ln(0) nrep(125) group(cabinetcode) constraints(1 2 3) 
>     from(largestparty_east =-0.772611 party_seatshare_east =0.0996268 largestparty=1.027553 party_seatshare=0.1369203);

Iteration 0:   log likelihood = -517.37549  (not concave)
Iteration 1:   log likelihood = -386.04625  (not concave)
Iteration 2:   log likelihood = -377.55725  (not concave)
Iteration 3:   log likelihood = -377.36439  
Iteration 4:   log likelihood = -375.50787  
Iteration 5:   log likelihood = -373.87861  
Iteration 6:   log likelihood = -372.93415  
Iteration 7:   log likelihood = -372.79813  
Iteration 8:   log likelihood = -372.75818  
Iteration 9:   log likelihood = -372.75676  
Iteration 10:  log likelihood = -372.75675  

Mixed logit model                                 Number of obs   =       2412
                                                  Wald chi2(2)    =       0.45
Log likelihood = -372.75675                       Prob > chi2     =     0.7986

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   1.687456   3.715275     0.45   0.650     -5.59435    8.969262
party_se~ast |  (omitted)
largestparty |   8.388737   12.77502     0.66   0.511    -16.64984    33.42731
party_seat~e |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |  -11.01092   21.26005    -0.52   0.605    -52.67986    30.65802
party_seat~e |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(3)  =    144.59
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

.       *     ****************************************************************  *;
. *       Model 1: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west,  
>     rand (largestparty party_seatshare) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -334.08276  (not concave)
Iteration 1:   log likelihood = -304.05869  
Iteration 2:   log likelihood = -303.96494  
Iteration 3:   log likelihood = -301.23316  
Iteration 4:   log likelihood =  -300.5095  
Iteration 5:   log likelihood = -300.46403  
Iteration 6:   log likelihood =  -300.4633  
Iteration 7:   log likelihood =  -300.4633  

Mixed logit model                                 Number of obs   =       2412
                                                  LR chi2(2)      =       8.00
Log likelihood =  -300.4633                       Prob > chi2     =     0.0183

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   .0772611   .8494944     0.09   0.928    -1.587717    1.742239
party_se~est |  -.0996268   .0634406    -1.57   0.116    -.2239681    .0247145
largestparty |   .9502914   .8692396     1.09   0.274    -.7533869     2.65397
party_seat~e |   .2365472   .0653712     3.62   0.000      .108422    .3646723
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.817643   1.078136    -2.61   0.009     -4.93075   -.7045348
party_seat~e |   -.008834   .0557698    -0.16   0.874    -.1181408    .1004727
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     *     ****************************************************************  *;
. *       Model 2: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east,  
>     rand (largestparty party_seatshare medianparty1) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -306.26012  (not concave)
Iteration 1:   log likelihood =  -277.5738  
Iteration 2:   log likelihood = -275.47476  
Iteration 3:   log likelihood = -275.40562  
Iteration 4:   log likelihood =  -275.3843  
Iteration 5:   log likelihood = -275.38394  
Iteration 6:   log likelihood = -275.38394  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(3)      =       5.10
Log likelihood = -275.38394                       Prob > chi2     =     0.1644

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.070898   .9471385    -1.13   0.258    -2.927256    .7854589
party_se~ast |   .1923742   .0856347     2.25   0.025     .0245333    .3602151
medianpa~ast |  -1.452603   .8286761    -1.75   0.080    -3.076779    .1715719
largestparty |   .7065556   .3552869     1.99   0.047     .0102061    1.402905
party_seat~e |   .1172551    .016713     7.02   0.000     .0844983     .150012
medianparty1 |   .9396789   .2245193     4.19   0.000     .4996292    1.379729
-------------+----------------------------------------------------------------
SD           |
largestparty |  -1.992631   .7200942    -2.77   0.006    -3.403989    -.581272
party_seat~e |  -.0053562   .0374201    -0.14   0.886    -.0786981    .0679858
medianparty1 |   .0123652   .9997304     0.01   0.990     -1.94707    1.971801
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. constraint 4 [SD]medianparty1 = 0;

. constraint 5 medianparty1 = 0;

. constraint 6 medianparty1_east = 0;

. mixlogit choice largestparty_east party_seatshare_east medianparty1_east,  
>     rand (largestparty party_seatshare medianparty1) ln(0) nrep(125) group(cabinetcode) constraints(1 2 3 4 5 6) 
>     from(largestparty_east =-1.070898 party_seatshare_east =0.1923742 medianparty1_east=-1.452603  largestparty=0.7065556 party_seatshare=0.1172551 medianparty1=0.9396789);

Iteration 0:   log likelihood = -537.27698  (not concave)
Iteration 1:   log likelihood = -376.39886  (not concave)
Iteration 2:   log likelihood = -365.39755  (not concave)
Iteration 3:   log likelihood = -365.27764  
Iteration 4:   log likelihood = -361.46261  
Iteration 5:   log likelihood = -360.21845  
Iteration 6:   log likelihood = -359.25993  
Iteration 7:   log likelihood = -358.34141  
Iteration 8:   log likelihood = -358.24381  
Iteration 9:   log likelihood = -358.20183  
Iteration 10:  log likelihood = -358.19947  
Iteration 11:  log likelihood = -358.19926  
Iteration 12:  log likelihood = -358.19925  

Mixed logit model                                 Number of obs   =       2269
                                                  Wald chi2(2)    =       0.04
Log likelihood = -358.19925                       Prob > chi2     =     0.9788

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
 ( 4)  [SD]medianparty1 = 0
 ( 5)  [Mean]medianparty1 = 0
 ( 6)  [Mean]medianparty1_east = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   .5886958   4.683167     0.13   0.900    -8.590143    9.767534
party_se~ast |  (omitted)
medianpa~ast |  (omitted)
largestparty |   11.71454   57.07753     0.21   0.837    -100.1554    123.5844
party_seat~e |  (omitted)
medianparty1 |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |  -16.54017   94.69099    -0.17   0.861    -202.1311    169.0507
party_seat~e |  (omitted)
medianparty1 |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(6)  =    165.63
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

. *     ****************************************************************  *;
. *       Model 2: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west medianparty1_west,  
>     rand (largestparty party_seatshare medianparty1) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -306.26012  (not concave)
Iteration 1:   log likelihood = -277.59582  
Iteration 2:   log likelihood = -275.48584  
Iteration 3:   log likelihood = -275.41063  
Iteration 4:   log likelihood = -275.38453  
Iteration 5:   log likelihood = -275.38394  
Iteration 6:   log likelihood = -275.38394  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(3)      =       5.10
Log likelihood = -275.38394                       Prob > chi2     =     0.1644

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   1.070898   .9471385     1.13   0.258    -.7854589    2.927256
party_se~est |  -.1923742   .0856347    -2.25   0.025    -.3602151   -.0245333
medianpa~est |   1.452603   .8286761     1.75   0.080    -.1715718    3.076779
largestparty |  -.3643428    .881452    -0.41   0.679    -2.091957    1.363271
party_seat~e |   .3096293   .0867622     3.57   0.000     .1395785    .4796802
medianparty1 |  -.5129246   .8013388    -0.64   0.522     -2.08352    1.057671
-------------+----------------------------------------------------------------
SD           |
largestparty |  -1.992631   .7200941    -2.77   0.006    -3.403989   -.5812721
party_seat~e |  -.0053561     .03742    -0.14   0.886    -.0786981    .0679858
medianparty1 |   .0123653   .9997302     0.01   0.990     -1.94707    1.971801
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

. *     ****************************************************************  *;
. *       Model 3: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -305.70032  (not concave)
Iteration 1:   log likelihood = -277.01128  
Iteration 2:   log likelihood = -272.93182  
Iteration 3:   log likelihood = -272.62239  
Iteration 4:   log likelihood = -272.60843  
Iteration 5:   log likelihood = -272.60839  
Iteration 6:   log likelihood = -272.60839  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(4)      =       8.75
Log likelihood = -272.60839                       Prob > chi2     =     0.0677

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.225344   1.117966    -1.10   0.273    -3.416517     .965828
party_se~ast |   .2197665   .0985857     2.23   0.026     .0265421    .4129909
medianpa~ast |  -1.380664   .9185156    -1.50   0.133    -3.180922    .4195933
presi~y_east |    .280208   1.300083     0.22   0.829    -2.267908    2.828324
largestparty |   .9123011   .4656598     1.96   0.050    -.0003753    1.824978
party_seat~e |   .1236354   .0181872     6.80   0.000     .0879893    .1592816
medianparty1 |   1.004963   .2543211     3.95   0.000     .5065028    1.503423
presidentp~y |   .6269142   .4810491     1.30   0.192    -.3159247    1.569753
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.652131   .9413996    -2.82   0.005     -4.49724   -.8070213
party_seat~e |   -.002107   .0342805    -0.06   0.951    -.0692955    .0650816
medianparty1 |   .1050148    .973465     0.11   0.914    -1.802941    2.012971
presidentp~y |   1.418201   .9582919     1.48   0.139    -.4600168    3.296418
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. constraint 4 [SD]medianparty1 = 0;

. constraint 5 medianparty1 = 0;

. constraint 6 medianparty1_east = 0;

. constraint 7 [SD]presidentparty = 0;

. constraint 8 presidentparty = 0;

. constraint 9 presidentparty_east = 0;

. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty) ln(0) nrep(125) group(cabinetcode) constraints(1 2 3 4 5 6 7 8 9) 
>     from(largestparty_east =-1.225344 party_seatshare_east=0.2197665 medianparty1_east=-1.380664  presidentparty_east=0.280208 
>     largestparty=0.9123011 party_seatshare=0.1236354 medianparty1=1.004963 presidentparty=0.6269142);

Iteration 0:   log likelihood = -510.55697  (not concave)
Iteration 1:   log likelihood = -377.23643  (not concave)
Iteration 2:   log likelihood = -365.41528  (not concave)
Iteration 3:   log likelihood =  -363.5553  
Iteration 4:   log likelihood = -360.45058  
Iteration 5:   log likelihood = -358.81533  
Iteration 6:   log likelihood = -358.33071  
Iteration 7:   log likelihood = -358.21407  
Iteration 8:   log likelihood = -358.20154  
Iteration 9:   log likelihood = -358.19933  
Iteration 10:  log likelihood = -358.19925  
Iteration 11:  log likelihood = -358.19925  

Mixed logit model                                 Number of obs   =       2269
                                                  Wald chi2(2)    =       0.04
Log likelihood = -358.19925                       Prob > chi2     =     0.9786

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
 ( 4)  [SD]medianparty1 = 0
 ( 5)  [Mean]medianparty1 = 0
 ( 6)  [Mean]medianparty1_east = 0
 ( 7)  [SD]presidentparty = 0
 ( 8)  [Mean]presidentparty = 0
 ( 9)  [Mean]presidentparty_east = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   .5881149   4.668641     0.13   0.900    -8.562253    9.738483
party_se~ast |  (omitted)
medianpa~ast |  (omitted)
presi~y_east |  (omitted)
largestparty |   11.70481   56.78313     0.21   0.837    -99.58809    122.9977
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |  -16.52413   94.20412    -0.18   0.861    -201.1608    168.1126
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(9)  =    171.18
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

. *     ****************************************************************  *;
. *       Model 3: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west medianparty1_west presidentparty_west,  
>     rand (largestparty party_seatshare medianparty1 presidentparty) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -305.70032  (not concave)
Iteration 1:   log likelihood = -277.23453  
Iteration 2:   log likelihood = -272.98915  
Iteration 3:   log likelihood = -272.64136  
Iteration 4:   log likelihood = -272.60843  
Iteration 5:   log likelihood = -272.60839  
Iteration 6:   log likelihood = -272.60839  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(4)      =       8.75
Log likelihood = -272.60839                       Prob > chi2     =     0.0677

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   1.225344   1.117966     1.10   0.273     -.965828    3.416517
party_se~est |  -.2197665   .0985857    -2.23   0.026    -.4129909   -.0265421
medianpa~est |   1.380664   .9185156     1.50   0.133    -.4195933    3.180922
presi~y_west |   -.280208   1.300083    -0.22   0.829    -2.828324    2.267908
largestparty |  -.3130432   1.037699    -0.30   0.763    -2.346896    1.720809
party_seat~e |   .3434019    .100537     3.42   0.001      .146353    .5404509
medianparty1 |  -.3757012   .8881626    -0.42   0.672    -2.116468    1.365066
presidentp~y |   .9071222   1.262465     0.72   0.472    -1.567263    3.381507
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.652131   .9413995    -2.82   0.005     -4.49724   -.8070215
party_seat~e |   -.002107   .0342805    -0.06   0.951    -.0692955    .0650816
medianparty1 |   .1050148    .973465     0.11   0.914    -1.802941    2.012971
presidentp~y |   1.418201   .9582919     1.48   0.139    -.4600168    3.296418
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     *     ****************************************************************  *;
. *       Model 4: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -302.23825  (not concave)
Iteration 1:   log likelihood =   -273.143  (not concave)
Iteration 2:   log likelihood = -273.08173  (not concave)
Iteration 3:   log likelihood = -271.49304  
Iteration 4:   log likelihood = -267.57901  (not concave)
Iteration 5:   log likelihood = -267.11521  
Iteration 6:   log likelihood = -266.83086  
Iteration 7:   log likelihood = -266.82235  
Iteration 8:   log likelihood = -266.82234  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(5)      =      13.09
Log likelihood = -266.82234                       Prob > chi2     =     0.0226

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.211664   1.116793    -1.08   0.278    -3.400539    .9772103
party_se~ast |    .224388   .0977387     2.30   0.022     .0328237    .4159523
medianpa~ast |  -1.517119   .9503912    -1.60   0.110    -3.379852    .3456134
presi~y_east |   .3653009   1.286935     0.28   0.777    -2.157045    2.887647
previ~M_east |  -1.358147   1.087967    -1.25   0.212    -3.490523    .7742293
largestparty |   .8590721   .4520203     1.90   0.057    -.0268715    1.745016
party_seat~e |   .1171662   .0193652     6.05   0.000     .0792112    .1551213
medianparty1 |   1.072249   .3000173     3.57   0.000     .4842258    1.660272
presidentp~y |   .5785964    .494909     1.17   0.242    -.3914074      1.5486
  previousPM |   .6955045    .350025     1.99   0.047     .0094681    1.381541
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.042723   1.159172    -1.76   0.078    -4.314658    .2292124
party_seat~e |  -.0026451   .0306948    -0.09   0.931    -.0628058    .0575156
medianparty1 |   -.370209   .9982212    -0.37   0.711    -2.326687    1.586268
presidentp~y |   1.274164   1.167069     1.09   0.275    -1.013249    3.561578
  previousPM |   -2.11954   1.030276    -2.06   0.040    -4.138845   -.1002352
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. constraint 4 [SD]medianparty1 = 0;

. constraint 5 medianparty1 = 0;

. constraint 6 medianparty1_east = 0;

. constraint 7 [SD]presidentparty = 0;

. constraint 8 presidentparty = 0;

. constraint 9 presidentparty_east = 0;

. constraint 10 [SD]previousPM = 0;

. constraint 11 previousPM = 0;

. constraint 12 previousPM_east = 0;

. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM) ln(0) nrep(125) group(cabinetcode) constraints(1 2 3 4 5 6 7 8 9 10 11 12) 
>     from(largestparty_east =-1.211664 party_seatshare_east=0.224388 medianparty1_east=-1.17119  presidentparty_east=0.3653009 previousPM_east=-1.358147 
>     largestparty=0.8590721 party_seatshare=0.1171662 medianparty1=1.072249 presidentparty=0.5785964 previousPM=0.6955045);

Iteration 0:   log likelihood =   -518.019  (not concave)
Iteration 1:   log likelihood = -377.23022  (not concave)
Iteration 2:   log likelihood = -365.56543  (not concave)
Iteration 3:   log likelihood = -364.35575  
Iteration 4:   log likelihood = -359.91089  
Iteration 5:   log likelihood = -358.87778  
Iteration 6:   log likelihood = -358.29869  
Iteration 7:   log likelihood = -358.24906  
Iteration 8:   log likelihood = -358.20488  
Iteration 9:   log likelihood = -358.19992  
Iteration 10:  log likelihood = -358.19925  
Iteration 11:  log likelihood = -358.19925  

Mixed logit model                                 Number of obs   =       2269
                                                  Wald chi2(2)    =       0.04
Log likelihood = -358.19925                       Prob > chi2     =     0.9787

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
 ( 4)  [SD]medianparty1 = 0
 ( 5)  [Mean]medianparty1 = 0
 ( 6)  [Mean]medianparty1_east = 0
 ( 7)  [SD]presidentparty = 0
 ( 8)  [Mean]presidentparty = 0
 ( 9)  [Mean]presidentparty_east = 0
 (10)  [SD]previousPM = 0
 (11)  [Mean]previousPM = 0
 (12)  [Mean]previousPM_east = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   .5882722   4.672672     0.13   0.900    -8.569997    9.746542
party_se~ast |  (omitted)
medianpa~ast |  (omitted)
presi~y_east |  (omitted)
previ~M_east |  (omitted)
largestparty |   11.70757   56.86562     0.21   0.837    -99.74701    123.1621
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |  -16.52871   94.34088    -0.18   0.861    -201.4334     168.376
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(12) =    182.75
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

. *     ****************************************************************  *;
. *       Model 4: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west medianparty1_west presidentparty_west previousPM_west,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -302.23825  (not concave)
Iteration 1:   log likelihood = -273.14036  (not concave)
Iteration 2:   log likelihood = -272.99307  (not concave)
Iteration 3:   log likelihood = -271.08768  (not concave)
Iteration 4:   log likelihood = -269.92729  
Iteration 5:   log likelihood = -267.03946  
Iteration 6:   log likelihood = -266.82919  
Iteration 7:   log likelihood = -266.82237  
Iteration 8:   log likelihood = -266.82234  
Iteration 9:   log likelihood = -266.82234  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(5)      =      13.09
Log likelihood = -266.82234                       Prob > chi2     =     0.0226

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   1.211666   1.116796     1.08   0.278    -.9772129    3.400545
party_se~est |  -.2243883   .0977389    -2.30   0.022    -.4159529   -.0328236
medianpa~est |   1.517119   .9503923     1.60   0.110    -.3456156    3.379854
presi~y_west |  -.3653022   1.286935    -0.28   0.777    -2.887648    2.157044
previ~M_west |    1.35815    1.08797     1.25   0.212    -.7742313    3.490532
largestparty |  -.3525916   1.021578    -0.35   0.730    -2.354847    1.649664
party_seat~e |   .3415545   .0997072     3.43   0.001     .1461321     .536977
medianparty1 |  -.4448698   .9001604    -0.49   0.621    -2.209152    1.319412
presidentp~y |   .9439032   1.253793     0.75   0.452    -1.513486    3.401292
  previousPM |  -.6626445   1.007298    -0.66   0.511    -2.636913    1.311623
-------------+----------------------------------------------------------------
SD           |
largestparty |  -2.042733   1.159185    -1.76   0.078    -4.314694     .229228
party_seat~e |   -.002645   .0306949    -0.09   0.931    -.0628059    .0575159
medianparty1 |  -.3701985   .9982377    -0.37   0.711    -2.326709    1.586312
presidentp~y |    1.27415     1.1671     1.09   0.275    -1.013324    3.561623
  previousPM |  -2.119543   1.030282    -2.06   0.040    -4.138859   -.1002271
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     *     ****************************************************************  *;
. *       Model 5: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east,   
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -299.01312  (not concave)
Iteration 1:   log likelihood = -269.99377  (not concave)
Iteration 2:   log likelihood = -269.96594  (not concave)
Iteration 3:   log likelihood = -269.16599  (not concave)
Iteration 4:   log likelihood = -268.51664  (not concave)
Iteration 5:   log likelihood = -267.02702  
Iteration 6:   log likelihood =  -264.8059  
Iteration 7:   log likelihood = -264.47127  
Iteration 8:   log likelihood = -264.46326  
Iteration 9:   log likelihood = -264.46313  
Iteration 10:  log likelihood = -264.46313  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(7)      =      11.08
Log likelihood = -264.46313                       Prob > chi2     =     0.1353

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.403309    1.19546    -1.17   0.240    -3.746369      .93975
party_se~ast |   .2506177    .106778     2.35   0.019     .0413366    .4598989
medianpa~ast |  -1.441322   .9633059    -1.50   0.135    -3.329367    .4467233
presi~y_east |   .9270295   1.373326     0.68   0.500    -1.764641    3.618699
previ~M_east |  -1.228084   1.128155    -1.09   0.276    -3.439227    .9830602
largestparty |   1.002409   .5148739     1.95   0.052    -.0067249    2.011544
party_seat~e |   .1165616   .0193914     6.01   0.000     .0785552     .154568
medianparty1 |   1.081568   .2915869     3.71   0.000     .5100684    1.653068
presidentp~y |   3.408346   1.502893     2.27   0.023       .46273    6.353962
  previousPM |   .7209862   .3752296     1.92   0.055    -.0144503    1.456423
president_~e |  -2.302214   1.200227    -1.92   0.055    -4.654616    .0501876
president~ct |  -2.332541   1.297614    -1.80   0.072    -4.875817    .2107356
-------------+----------------------------------------------------------------
SD           |
largestparty |   2.500089   1.180402     2.12   0.034     .1865429    4.813634
party_seat~e |  -.0012578   .0285543    -0.04   0.965    -.0572231    .0547075
medianparty1 |   .1329975   .9370381     0.14   0.887    -1.703563    1.969559
presidentp~y |    .089412   1.333708     0.07   0.947    -2.524607    2.703431
  previousPM |   1.914019   .8628325     2.22   0.027     .2228983     3.60514
president_~e |  -.1686345   2.643274    -0.06   0.949    -5.349356    5.012087
president~ct |     .52211   1.409805     0.37   0.711    -2.241056    3.285276
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.        *     ****************************************************************  *;
. *       Model 5 for Western Europe reaches a local maximum and so we    *;
. *       need to give it better starting values.  We use the starting    *;
. *       values based on the results from Model 5 for Western Europe     *;
. *       except that we use the coefficient estimates on the president   *;
. *       interaction term from Model 5 for Eastern Europe.               *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east,    
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct) ln(0) nrep(125) group(cabinetcode) 
>     from(largestparty_east =-1.403309 party_seatshare_east =0.2506177 medianparty1_east =-1.441322 presidentparty_east =0.9270295 
>     previousPM_east =-1.228084 largestparty=1.002409 party_seatshare=0.1165616 medianparty1=1.081568 presidentparty=3.408346 
>     previousPM=0.7209862 president_investiture= -2.267492 president_direct= -2.237836);

Iteration 0:   log likelihood = -299.32254  (not concave)
Iteration 1:   log likelihood = -272.22809  (not concave)
Iteration 2:   log likelihood = -272.09903  (not concave)
Iteration 3:   log likelihood = -271.23613  (not concave)
Iteration 4:   log likelihood = -269.80427  (not concave)
Iteration 5:   log likelihood = -267.19917  (not concave)
Iteration 6:   log likelihood = -265.75866  (not concave)
Iteration 7:   log likelihood = -264.89425  
Iteration 8:   log likelihood = -264.44142  
Iteration 9:   log likelihood = -264.42693  
Iteration 10:  log likelihood = -264.42686  
Iteration 11:  log likelihood = -264.42686  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(7)      =      11.15
Log likelihood = -264.42686                       Prob > chi2     =     0.1322

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.359312   1.144521    -1.19   0.235    -3.602533    .8839087
party_se~ast |   .2431044   .1025968     2.37   0.018     .0420183    .4441905
medianpa~ast |  -1.439561   .9452127    -1.52   0.128    -3.292144    .4130219
presi~y_east |   .9508789   1.322865     0.72   0.472    -1.641889    3.543647
previ~M_east |  -1.108167   1.089611    -1.02   0.309    -3.243765    1.027431
largestparty |   .9542042   .4757461     2.01   0.045     .0217591    1.886649
party_seat~e |   .1153319   .0188941     6.10   0.000     .0783002    .1523636
medianparty1 |   1.084636   .2914065     3.72   0.000     .5134896    1.655782
presidentp~y |   3.318009   1.403304     2.36   0.018     .5675841    6.068434
  previousPM |   .6825148   .3616772     1.89   0.059    -.0263595    1.391389
president_~e |  -2.267492   1.147815    -1.98   0.048    -4.517169   -.0178156
president~ct |  -2.237836   1.227987    -1.82   0.068    -4.644646    .1689736
-------------+----------------------------------------------------------------
SD           |
largestparty |   2.269877   1.092067     2.08   0.038     .1294649    4.410288
party_seat~e |   .0009964   .0308224     0.03   0.974    -.0594144    .0614071
medianparty1 |   .0416482   .9169799     0.05   0.964    -1.755599    1.838896
presidentp~y |   .4045558   1.396475     0.29   0.772    -2.332484    3.141596
  previousPM |   -1.99265   .9469771    -2.10   0.035    -3.848691   -.1366094
president_~e |   .2151218   1.895196     0.11   0.910    -3.499394    3.929638
president~ct |  -.1541027   1.807404    -0.09   0.932     -3.69655    3.388344
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. constraint 4 [SD]medianparty1 = 0;

. constraint 5 medianparty1 = 0;

. constraint 6 medianparty1_east = 0;

. constraint 7 [SD]presidentparty = 0;

. constraint 8 presidentparty = 0;

. constraint 9 presidentparty_east = 0;

. constraint 10 [SD]previousPM = 0;

. constraint 11 previousPM = 0;

. constraint 12 previousPM_east = 0;

. constraint 13 [SD]president_investiture = 0;

. constraint 14 president_investiture = 0;

. constraint 15 [SD]president_direct = 0;

. constraint 16 president_direct = 0;

. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east,   
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct) ln(0) nrep(125) group(cabinetcode) 
>     constraints(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16) 
>     from(largestparty_east =-1.403309 party_seatshare_east =0.2506177 medianparty1_east =-1.441322 presidentparty_east =0.9270295 
>     previousPM_east =-1.228084 largestparty=1.002409 party_seatshare=0.1165616 medianparty1=1.081568 presidentparty=3.408346 
>     previousPM=0.7209862 president_investiture= -2.267492 president_direct= -2.237836);

Iteration 0:   log likelihood = -503.08198  (not concave)
Iteration 1:   log likelihood =  -379.8618  (not concave)
Iteration 2:   log likelihood = -364.52318  
Iteration 3:   log likelihood = -360.34217  
Iteration 4:   log likelihood = -358.58087  
Iteration 5:   log likelihood = -358.29673  
Iteration 6:   log likelihood = -358.20312  
Iteration 7:   log likelihood = -358.19963  
Iteration 8:   log likelihood = -358.19926  
Iteration 9:   log likelihood = -358.19925  

Mixed logit model                                 Number of obs   =       2269
                                                  Wald chi2(2)    =       0.04
Log likelihood = -358.19925                       Prob > chi2     =     0.9785

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
 ( 4)  [SD]medianparty1 = 0
 ( 5)  [Mean]medianparty1 = 0
 ( 6)  [Mean]medianparty1_east = 0
 ( 7)  [SD]presidentparty = 0
 ( 8)  [Mean]presidentparty = 0
 ( 9)  [Mean]presidentparty_east = 0
 (10)  [SD]previousPM = 0
 (11)  [Mean]previousPM = 0
 (12)  [Mean]previousPM_east = 0
 (13)  [SD]president_investiture = 0
 (14)  [Mean]president_investiture = 0
 (15)  [SD]president_direct = 0
 (16)  [Mean]president_direct = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   .5877663   4.659461     0.13   0.900     -8.54461    9.720143
party_se~ast |  (omitted)
medianpa~ast |  (omitted)
presi~y_east |  (omitted)
previ~M_east |  (omitted)
largestparty |   11.69839   56.59318     0.21   0.836     -99.2222     122.619
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
president_~e |  (omitted)
president~ct |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |  -16.51342   93.88867    -0.18   0.860    -200.5318     167.505
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
president_~e |  (omitted)
president~ct |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(16) =    187.54
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

. *     ****************************************************************  *;
. *       Model 5: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west medianparty1_west presidentparty_west previousPM_west,   
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct) ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -299.01312  (not concave)
Iteration 1:   log likelihood = -269.99767  (not concave)
Iteration 2:   log likelihood = -269.95646  (not concave)
Iteration 3:   log likelihood = -269.89412  (not concave)
Iteration 4:   log likelihood = -267.51782  (not concave)
Iteration 5:   log likelihood = -267.09521  
Iteration 6:   log likelihood = -264.86743  
Iteration 7:   log likelihood =  -264.4438  
Iteration 8:   log likelihood = -264.42693  
Iteration 9:   log likelihood = -264.42686  
Iteration 10:  log likelihood = -264.42686  

Mixed logit model                                 Number of obs   =       2269
                                                  LR chi2(7)      =      11.15
Log likelihood = -264.42686                       Prob > chi2     =     0.1322

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   1.359312   1.144521     1.19   0.235    -.8839087    3.602533
party_se~est |  -.2431044   .1025968    -2.37   0.018    -.4441905   -.0420183
medianpa~est |   1.439561   .9452127     1.52   0.128    -.4130219    3.292144
presi~y_west |  -.9508789   1.322865    -0.72   0.472    -3.543647    1.641889
previ~M_west |   1.108167   1.089611     1.02   0.309    -1.027431    3.243765
largestparty |   -.405108   1.041478    -0.39   0.697    -2.446367    1.636152
party_seat~e |   .3584363   .1049154     3.42   0.001     .1528058    .5640668
medianparty1 |  -.3549251   .8999642    -0.39   0.693    -2.118823    1.408972
presidentp~y |   4.268888   1.880169     2.27   0.023     .5838251    7.953951
  previousPM |  -.4256523   .9966841    -0.43   0.669    -2.379117    1.527813
president_~e |  -2.267492   1.147815    -1.98   0.048    -4.517169   -.0178156
president~ct |  -2.237836   1.227987    -1.82   0.068    -4.644646    .1689735
-------------+----------------------------------------------------------------
SD           |
largestparty |   2.269877   1.092067     2.08   0.038     .1294651    4.410288
party_seat~e |   .0009964   .0308224     0.03   0.974    -.0594144    .0614071
medianparty1 |   .0416482   .9169799     0.05   0.964    -1.755599    1.838896
presidentp~y |   .4045558   1.396475     0.29   0.772    -2.332484    3.141596
  previousPM |   -1.99265   .9469771    -2.10   0.035    -3.848691   -.1366094
president_~e |   .2151218   1.895196     0.11   0.910    -3.499394    3.929638
president~ct |  -.1541028   1.807404    -0.09   0.932     -3.69655    3.388344
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     *     ****************************************************************  *;
. *       Model 6: Western Europe (with LR test)                          *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east previousPM_conflict_east previousPM_gain1_east previous_cabinet_gain1_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct previousPM_conflict previousPM_gain1 previous_cabinet_gain1) 
>     ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -255.29786  (not concave)
Iteration 1:   log likelihood = -245.38797  (not concave)
Iteration 2:   log likelihood =  -228.8695  
Iteration 3:   log likelihood = -221.24204  (not concave)
Iteration 4:   log likelihood = -221.16184  (not concave)
Iteration 5:   log likelihood = -220.39987  
Iteration 6:   log likelihood = -219.61402  
Iteration 7:   log likelihood = -219.46931  
Iteration 8:   log likelihood = -219.42816  (not concave)
Iteration 9:   log likelihood = -219.41127  (not concave)
Iteration 10:  log likelihood = -219.40734  (not concave)
Iteration 11:  log likelihood = -219.38014  
Iteration 12:  log likelihood = -219.35916  
Iteration 13:  log likelihood = -219.35268  
Iteration 14:  log likelihood = -219.35266  

Mixed logit model                                 Number of obs   =       2039
                                                  LR chi2(10)     =      19.54
Log likelihood = -219.35266                       Prob > chi2     =     0.0339

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |  -1.253544   2.017873    -0.62   0.534    -5.208503    2.701414
party_se~ast |   .2877481   .1444587     1.99   0.046     .0046143    .5708819
medianpa~ast |  -3.117227   1.721757    -1.81   0.070    -6.491809    .2573546
presi~y_east |  -.4970681   1.853252    -0.27   0.789    -4.129376     3.13524
previ~M_east |   -2.19965   2.606521    -0.84   0.399    -7.308337    2.909037
previ~t_east |   1.126725   3.208478     0.35   0.725    -5.161776    7.415226
previousPM.. |  -.0930831   .4810077    -0.19   0.847    -1.035841    .8496747
previous_c.. |   .0517841   .1988117     0.26   0.795    -.3378797    .4414479
largestparty |   2.109019   1.104417     1.91   0.056    -.0555996    4.273637
party_seat~e |   .1351541   .0258657     5.23   0.000     .0844583    .1858499
medianparty1 |   1.392319   .3828235     3.64   0.000     .6419985    2.142639
presidentp~y |   5.250695   2.428888     2.16   0.031      .490162    10.01123
  previousPM |   2.841572   1.038968     2.73   0.006     .8052325    4.877912
president_~e |  -3.060496   2.130412    -1.44   0.151    -7.236027    1.115035
president~ct |  -3.915977     2.1442    -1.83   0.068    -8.118531    .2865775
previousP~ct |  -2.618633   1.261868    -2.08   0.038    -5.091848   -.1454174
previousPM~1 |   .4958572   .2007184     2.47   0.013     .1024564     .889258
previous_c~1 |   .0367127   .0403208     0.91   0.363    -.0423147    .1157401
-------------+----------------------------------------------------------------
SD           |
largestparty |   -4.72622   1.779758    -2.66   0.008    -8.214482   -1.237958
party_seat~e |  -.0074871   .0352742    -0.21   0.832    -.0766233    .0616491
medianparty1 |    .137813   .9798871     0.14   0.888     -1.78273    2.058357
presidentp~y |   -.266903   1.275177    -0.21   0.834    -2.766204    2.232398
  previousPM |   -1.90488   1.048407    -1.82   0.069     -3.95972    .1499603
president_~e |  -.4246113   1.592647    -0.27   0.790    -3.546143     2.69692
president~ct |   .5917079    1.45034     0.41   0.683    -2.250906    3.434321
previousP~ct |  -.1339581   2.773063    -0.05   0.961    -5.569061    5.301145
previousPM~1 |     .11201   .2052589     0.55   0.585      -.29029      .51431
previous_c~1 |   .1484155   .0862285     1.72   0.085    -.0205892    .3174201
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

. estimates store A;

. constraint 1 [SD]party_seatshare = 0;

. constraint 2 party_seatshare = 0;

. constraint 3 party_seatshare_east = 0;

. constraint 4 [SD]medianparty1 = 0;

. constraint 5 medianparty1 = 0;

. constraint 6 medianparty1_east = 0;

. constraint 7 [SD]presidentparty = 0;

. constraint 8 presidentparty = 0;

. constraint 9 presidentparty_east = 0;

. constraint 10 [SD]previousPM = 0;

. constraint 11 previousPM = 0;

. constraint 12 previousPM_east = 0;

. constraint 13 [SD]president_investiture = 0;

. constraint 14 president_investiture = 0;

. constraint 15 [SD]president_direct = 0;

. constraint 16 president_direct = 0;

. constraint 17 [SD]previousPM_conflict = 0;

. constraint 18 previousPM_conflict = 0;

. constraint 19 previousPM_conflict_east = 0;

. constraint 20 [SD]previousPM_gain1 = 0;

. constraint 21 previousPM_gain1 = 0;

. constraint 22 previousPM_gain1_east = 0;

. constraint 23 [SD]previous_cabinet_gain1 = 0;

. constraint 24 previous_cabinet_gain1 = 0;

. constraint 25 previous_cabinet_gain1_east = 0;

. mixlogit choice largestparty_east party_seatshare_east medianparty1_east presidentparty_east previousPM_east previousPM_conflict_east previousPM_gain1_east previous_cabinet_gain1_east,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct previousPM_conflict previousPM_gain1 previous_cabinet_gain1) 
>     ln(0) nrep(125) group(cabinetcode) constraints(1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25) 
>     from(largestparty_east =-1.253544 party_seatshare_east=0.2877481 medianparty1_east =-3.117227 presidentparty_east =0.4970681 
>     previousPM_east =-2.19965 previousPM_conflict_east=1.126725 previousPM_gain1_east=-0.0930831 previous_cabinet_gain1_east=0.0517841
>     largestparty=2.109019 party_seatshare=0.1351541 medianparty1=1.392319 presidentparty=5.250695 
>     previousPM=2.841572 president_investiture= -3.060496 president_direct= -3.915977 previousPM_conflict=-2.618633 previousPM_gain1=0.4958572 previous_cabinet_gain1=0.0367127);

Iteration 0:   log likelihood = -342.92857  (not concave)
Iteration 1:   log likelihood =  -320.0615  (not concave)
Iteration 2:   log likelihood =  -319.1765  
Iteration 3:   log likelihood = -316.84197  
Iteration 4:   log likelihood = -315.41331  
Iteration 5:   log likelihood = -315.09579  
Iteration 6:   log likelihood = -314.87418  
Iteration 7:   log likelihood = -314.85806  
Iteration 8:   log likelihood = -314.85735  
Iteration 9:   log likelihood = -314.85735  

Mixed logit model                                 Number of obs   =       2039
                                                  Wald chi2(2)    =       0.30
Log likelihood = -314.85735                       Prob > chi2     =     0.8628

 ( 1)  [SD]party_seatshare = 0
 ( 2)  [Mean]party_seatshare = 0
 ( 3)  [Mean]party_seatshare_east = 0
 ( 4)  [SD]medianparty1 = 0
 ( 5)  [Mean]medianparty1 = 0
 ( 6)  [Mean]medianparty1_east = 0
 ( 7)  [SD]presidentparty = 0
 ( 8)  [Mean]presidentparty = 0
 ( 9)  [Mean]presidentparty_east = 0
 (10)  [SD]previousPM = 0
 (11)  [Mean]previousPM = 0
 (12)  [Mean]previousPM_east = 0
 (13)  [SD]president_investiture = 0
 (14)  [Mean]president_investiture = 0
 (15)  [SD]president_direct = 0
 (16)  [Mean]president_direct = 0
 (17)  [SD]previousPM_conflict = 0
 (18)  [Mean]previousPM_conflict = 0
 (19)  [Mean]previousPM_conflict_east = 0
 (20)  [SD]previousPM_gain1 = 0
 (21)  [Mean]previousPM_gain1 = 0
 (22)  [Mean]previousPM_gain1_east = 0
 (23)  [SD]previous_cabinet_gain1 = 0
 (24)  [Mean]previous_cabinet_gain1 = 0
 (25)  [Mean]previous_cabinet_gain1_east = 0
------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~ast |   -.088369   3.222315    -0.03   0.978     -6.40399    6.227252
party_se~ast |  (omitted)
medianpa~ast |  (omitted)
presi~y_east |  (omitted)
previ~M_east |  (omitted)
previ~t_east |  (omitted)
previousPM.. |  (omitted)
previous_c.. |  (omitted)
largestparty |   10.40884   19.18428     0.54   0.587    -27.19165    48.00934
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
president_~e |  (omitted)
president~ct |  (omitted)
previousP~ct |  (omitted)
previousPM~1 |  (omitted)
previous_c~1 |  (omitted)
-------------+----------------------------------------------------------------
SD           |
largestparty |   13.51232   29.75409     0.45   0.650    -44.80462    71.82926
party_seat~e |  (omitted)
medianparty1 |  (omitted)
presidentp~y |  (omitted)
  previousPM |  (omitted)
president_~e |  (omitted)
president~ct |  (omitted)
previousP~ct |  (omitted)
previousPM~1 |  (omitted)
previous_c~1 |  (omitted)
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

.     estimates store B;

. lrtest A B;

Likelihood-ratio test                                  LR chi2(25) =    191.01
(Assumption: B nested in A)                            Prob > chi2 =    0.0000

. *     ****************************************************************  *;
. *       Model 6: Eastern Europe                                         *;
. *     ****************************************************************  *;
. mixlogit choice largestparty_west party_seatshare_west medianparty1_west presidentparty_west previousPM_west previousPM_conflict_west previousPM_gain1_west previous_cabinet_gain1_west,  
>     rand (largestparty party_seatshare medianparty1 presidentparty previousPM president_investiture president_direct previousPM_conflict previousPM_gain1 previous_cabinet_gain1) 
>     ln(0) nrep(125) group(cabinetcode);

Iteration 0:   log likelihood = -255.29786  (not concave)
Iteration 1:   log likelihood = -244.91958  (not concave)
Iteration 2:   log likelihood =  -228.6947  
Iteration 3:   log likelihood = -222.07337  (not concave)
Iteration 4:   log likelihood = -221.27745  
Iteration 5:   log likelihood = -220.44537  (not concave)
Iteration 6:   log likelihood = -219.94673  (not concave)
Iteration 7:   log likelihood = -219.73862  
Iteration 8:   log likelihood = -219.42804  
Iteration 9:   log likelihood = -219.35695  
Iteration 10:  log likelihood = -219.35268  
Iteration 11:  log likelihood = -219.35266  

Mixed logit model                                 Number of obs   =       2039
                                                  LR chi2(10)     =      19.54
Log likelihood = -219.35266                       Prob > chi2     =     0.0339

------------------------------------------------------------------------------
      choice |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
Mean         |
largestp~est |   1.253544   2.017878     0.62   0.534    -2.701423    5.208512
party_se~est |  -.2877482   .1444587    -1.99   0.046    -.5708821   -.0046143
medianpa~est |   3.117229   1.721758     1.81   0.070    -.2573555    6.491813
presi~y_west |   .4970704   1.853254     0.27   0.789     -3.13524    4.129381
previ~M_west |   2.199652   2.606526     0.84   0.399    -2.909045    7.308349
previ~t_west |  -1.126728   3.208481    -0.35   0.725    -7.415236     5.16178
previousPM.. |    .093082   .4810118     0.19   0.847    -.8496838    1.035848
previous_c.. |  -.0517839   .1988119    -0.26   0.795    -.4414481    .3378803
largestparty |   .8554807   2.088361     0.41   0.682    -3.237632    4.948593
party_seat~e |   .4229023    .147261     2.87   0.004      .134276    .7115286
medianparty1 |   -1.72491   1.656514    -1.04   0.298    -4.971618    1.521798
presidentp~y |   4.753632   2.864886     1.66   0.097    -.8614405     10.3687
  previousPM |   .6419236   2.378381     0.27   0.787    -4.019618    5.303465
president_~e |  -3.060502   2.130416    -1.44   0.151    -7.236041    1.115036
president~ct |  -3.915981   2.144205    -1.83   0.068    -8.118545    .2865827
previousP~ct |   -1.49191   3.024037    -0.49   0.622    -7.418913    4.435094
previousPM~1 |   .4027744   .4782139     0.84   0.400    -.5345077    1.340056
previous_c~1 |   .0884968   .1965176     0.45   0.652    -.2966707    .4736642
-------------+----------------------------------------------------------------
SD           |
largestparty |  -4.726231   1.779764    -2.66   0.008    -8.214505   -1.237957
party_seat~e |  -.0074869   .0352743    -0.21   0.832    -.0766232    .0616494
medianparty1 |   .1378141   .9798873     0.14   0.888     -1.78273    2.058358
presidentp~y |  -.2668973   1.275176    -0.21   0.834    -2.766196    2.232402
  previousPM |  -1.904887    1.04841    -1.82   0.069    -3.959734    .1499596
president_~e |   -.424616   1.592646    -0.27   0.790    -3.546144    2.696912
president~ct |    .591713   1.450332     0.41   0.683    -2.250885    3.434311
previousP~ct |   -.133975   2.772999    -0.05   0.961    -5.568953    5.301003
previousPM~1 |    .112006    .205256     0.55   0.585    -.2902885    .5143004
previous_c~1 |   .1484154   .0862277     1.72   0.085    -.0205878    .3174186
------------------------------------------------------------------------------
The sign of the estimated standard deviations is irrelevant: interpret them as
being positive

. log close;
      name:  <unnamed>
       log:  C:\fsu2010\formateur\replication\stata_mxl.log
  log type:  text
 closed on:   5 Nov 2010, 16:33:37
---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
